stringfellow leachate treatment with rbc

Stringfellow Leachate Treatment with RBC

E. J. Opatken, H. K. Howard, and J. J. Bond U.S. Environmental Protection Agency, Cincinnati, Ohio 45268

A study was conducted with a rotating biological contactor (RBC) for treatment of leachate from the Stringfellow hazardous waste site in Riverside County, California. The leachate was transported from

Cal$ornia to Cincinnati, where a pilot sized RBC was installed at the U . S . EPA’s Testing and Evaluation ( T b E ) Facility.

A series of kinetic runs were made with primary effluent from the City of Cincinnati’s Mill Creek Sewage Treatment Plant to develop a biomass on

the disks and to obtain a standard kinetic removal rate. These runs were then followed with Stringfellow leachate experiments that included

Operations at various ratios of leachate to primary effluent Operations at 100% leachate Operations to increase the percentage removal of dissolved organics This paper reports on the results from these experiments and the

effectiveness of an RBC to adequately treat leachate from the Stringfellow hazardous waste site.

INTRODUCTION

Stringfellow is ii hazardons waste site that is located i n Glen Avon, California which is near Riverside, Califi)r- nia. The leachate is generated at an approxiniate rate of96 in’id (ZSOOO galid) and contains high concentrations of metals and organics. A leachate treatment fiacility oper- ates at the site aiid includes lime treatment for metals rc- moval, followed by clarification, sand filtration, and granular carllon treatment. The effluent froin the carl,on I)etls is then trucked 19 kin (12 mi) to an Orange County inter- ceptor sewer for disposal.

PROJECT DESCRIPTION

A pilot sized HBC was installed at the EPA Testing and Evaluation Facility (T&E) in Cincinnati for evaluating the biochemical treatability of Stringfellow leachate. The pilot sized HBC contains 1000 Inr (11000 sq ft) of snrface area which is approximately 10% ofa full-scale RBC. The dianieter of the pilot unit is 3.6 m (12 ft), which is identi- cal to a fiill scale RBC. The length is less than 1 111 (3.3 ft), whereas a fiill scale RBC is 7.6 m (25 ft). The Cincinnati treatment plant’s primary effluent (PE) was used to tle- velop a liiological popnlation and to obtain basic kinetic data. The ideal condition would have involved rising Stringfellow leachate to develop an indigenous bioma However, the logistics favored the treatment plants PE due to the vast distance between Cincinnati and River- side, CA. There were plus fiactors associated with con- dncting the experiment in Cincinnati, such a s the auxil- iary support available for installation, chemical analyses, 24 hour monitoring, and biochemical expertise for con- sultation. The RBC facility was designed to operate in a

Environmental Progress (Vol. 7, No. 1)

batch mode with the 5000 gallons of leachate that was trucked from California to Cincinnati for experimen- tation. The leachate was lime treated at Stringfellow for metals reduction so that the experiments could concen- trate on soln1,le organics removal with the RBC treatment. The experiments were designed to operate in a batch mode for the following reasons:

eliminate flow controls minimize spillages minimize accidental releases into the TCIrE sewer systems improve mass balance analyses obtain reaction kinetics data control final disposal avoid overloading of the shaft can be directly scaled for the Stringfellow Site

OBJECTIVE

The prime objective of this project was to determine whether the Stringfellow leachate can Iie economically converted into an innocuous waste by biochemical treatment with a rotating biological contactor.

Methodology

The RBC was operated with primary effluent supplied from the Metropolitan Sewer Districts of Greater Cincin- nati (MSDGC) Mill Creek Treatment Plant (MCTP) to develop an adequate biomass on the RBC disks in prepa- ration for leachate treatment.

The first experimental batches were made with increasing ratios of leachate/RBC treated primary effluent to

February, 1988 41

-

STORAGE TANK

TANK TRUCK

MILL CREEK PRIMARY EFFLUENl

ROTATING BIOLOGICAL CONTACTOR

Figure 1. Stringfellow-RBC flow schematic.

allow a gradual accliiriation period of the I)iomass to the leachate. Following these Iiatches the runs were made with 100% leachate. The operation consisted of transfer- ring the leachate from the storage tank to the mix tank where the voluine was determined. It was then pumped to the HBC as shown in Figure 1. The HBC was operated at a speed of 1.5 rpni and the operation continued mntil the dissolved organic carbon (DOC) reached a constant level. The HBC contents were then returned to the mix tank for additional treatment with activated powdered carbon (APC) followed by clarification with ferric chloride to achieve the effluent standards required for disposal to the MCTP. The effluent limitations set by the MSD were: Total organic halides (TOX) < 5 nig/L; vapor space organics (VSO) < 300 ppm; 6 < pH < 10.

Internal limits were a l s o set on gross organics that were equivalent to ii relatively high strength raw wastewater. These limits were: soluble biochcmical oxygen demand, SBOD < 100 mg/L; dissolved organic carbon, IIOC, < 100 mglL; soliible chemical oxygen demand, SCOD, < 300 mg/L.

During the experiment the following parameters were monitored or recorded: Volume of leachate in the HBC; temperature, every 2 hours; speed of rotation, every G hours; visual comments on thickness and color of bio- inass, every 2 hours; dissolved oxygen, every 2 hours; pH, every 2 hours (adjusted with soidriin hydroxide when necessary to maintain the pH above 6).

Kinetic Parameters

During the experiment, samples were obtained from the HBC tank at specified time intervals to determine the disappearance of the soluble gross organics. The soluble material is defined a s the liqriid phase that passes through a Whatman 934AH filter. The gross organics monitored incl~tded SBOD, DOC, and SCOD. During each exper-

42 February, 1988

Primary effluenc Run P2 3/5/66 2oh 100

2 4

Time. h

Figure 2. Disappearance of soluble gross organics with time.

Primary effluent after Batch 3 614186

1 DOC 400r SCOD

--- 2 4

Time, h

Figure 3. Disappearance of soluble gross organics with time.


TABLE 1. SOLUBLE CROSS ORGANICS REMOVAL FOR PE

Six Hour Residence Time

SBOD, mg/L DOC, mg/L SCOD, mg/L Date In Out In o u t In o u t - - - - - - - 6/20 6/19 6/18 6/16

77 99 72 55

0.94 0.00 0.8 1.3

iment the DOC was determined on site to obtain information on the progress of the reaction.

In addition to the gross organics, the raw leachate and the treated effluent were analyzed for: suspended solids, (SS); volatile suspended solids, (VSS); nitrogen series, in- cluding Kjeldahl, ammonia, nitrite, and nitrates phos- phorus, (P); and specific organic contaminants; para-chlorobenzene sulfonic acid (CBSA); 1,2 dichlorobenzene; o-xylene; chloroform; ethylbenzene; 2-hexanone; tetra- chloroeth ylene.

RESULTS AND DISCUSSION

The rate of soluble organic removals was determined for the Mill Creek primary effluent by following the disappearance of SBOD, SCOD and DOC with time. A typical curve for the disappearance of soluble organics using

Leachate I 330 gat 2 670 gal

Batch 1 - PE

40 I I I I ~ I I I I ~

2 4 6 8 10 12 14 16 18 20

Leachate Batch 2

3 I

= - = 750 gal. 25Ogaf.

100

Time, d Figure 4. Disappearance of DOC with time using various ratios of

leachate to treated effluent.


85 74 70 57

20 10 13 11

340 310 300 210

54 20 27 41

PE is displayed in Figure 2 and this curve is used as a base for comparison with PE following experiments with leachate on the RBC. The removal rate is typical for PE which can best be described as a high rate process as indicated by the rapid dropoff in gross organics after two hours of reaction time.

After three experiments with diluted leachate, there oc- curred a slough off of biomass from the disks as witnessed by an increase of solids in the RBC tank. It was question- able as to whether sufficient biomass was present on the disks, therefore the leachate was replaced with PE and an experiment was conducted to obtain the rate of disappearance of soluble organics with time. The curve in Fig- ure 3 represents the reduction of soluble organics with time for PE. The curve differs considerably from the rapid removal rate experienced with PE in Figure 2. It does not show the initial rapid drop-off of organics or the low residual organics that were previously obtained with PE, indicating that the RBC may have become biomass limited. To correct this situation the hiomass was rebuilt using primary effluent with batch times of six hours and again the removals returned to the rapid rate with a relatively low residual as shown in Table 1.

BIOMASS ACCLIMATION

Since the biomass established on the RBC disks was developed with primary effluent, acclimation runs were made to avoid shock loading the biomass with the Stringfellow leachate. The leachate was three to four times stronger in organic concentration than the PE and contained a significant quantity of p-chlorobenzene sui- fonic acid, a compound that is foreign to the biomass. The acclimation period consisted of three runs using progres- sively stronger concentrations of leachate as shown below. The raito of leachate to PE were:

Batch #1: Leachateltreated PE = 112 Batch #2: Leachateltreated batch #1 = 111 Batch #3: Leachateltreated batch #2 = 311

When leachate was combined with treated PE in batch #1 there were six passive days before the DOC began

Time. d

Figure 5. Disappearance of soluble organics with time.

February, 1988 43

to disappear. After the incubation period the reaction required approximately 2 days to obtain a drop ofapproxi- mately 60% in the DOC. The residual DOC, approximately 50 mg/L, remained unchanged during an additional 6 days of treatment. This is illustrated in Figure 4 in which the concentration of DOC is plotted against time for the first three batches.

In batch #2 the reaction for the disappearance of DOC began after the first day and was essentially complete after three days. The percent removal of DOC was approximately 50% and the residual DOC was approximately 90 mg/L. The result for batch #3 was similar to batch #2. That is, after a passive 24 hours, the reduction in DOC was essentially complete after three days. The DOC removal was approximately 55% and the batch had a final DOC of 120 mg/L.

Batch #3 was continued for an additional 27 days to determine if an extended time period would result in a further reduction in DOC. The additional time had no effect on DOC as shown in Figure 4.

100% LEACHATE OPERATION

Following the acclimation runs, where increasing concentrations of leachate were used to gradually expose the biomass to the leachate, the next two runs used 100% leachate.

The results were similar to those obtained during the acclimation runs 2 and 3. A plot of the soluble organics removal versus time for SBOD, DOC and SCOD is shown in Figure 5. We conclude from this data that: approximately four days were required to reduce the DOC from 300 to 100 mg/L. This closely approximates the results obtained with the leachate and PE mixture and is in con- trast to only 2 hours that is needed to obtain a large DOC reduction with PE.

The removal rate is significantly below the rate obtained with PE. Previous work on RBC’s showed a SBOD removal rate of 440 mg/h.m2 (40 mg/h.ft2) with municipal wastewater; whereas the leachate gave 19 mg/h.m2 (1.7 mg/h. ft’).

The RBC operation required pH adjustment. As the DOC dropped during the reaction phase the pH would fall and require periodic additions of sodium hydroxide to maintain a pH greater than six.

Sodium phosphate was added to the leachate to provide adequate levels of P to maintain a satisfactory supply of nutrients. Characterization of the leachate indicated that

TABLE 2. TREATMENT OF 100% STRINGFELLOW LEACHATE

Leachate mg/L

SBOD BOD DOC TOC SCOD COD ss vss TKN NH3-N N03-N NO*-N P

420 440 300 310 800 840 43 31 6.3 3.4

44 ND*

3.5

* Not detected.

44 February, 1988

APC + RBC Clarification

Effluent, Effluent mg/L mg/L

0.0 0.9 2.2

110 20 22

360 79 95 23 14 7.5 6.3

34 ND*

2.3

nitrogen was readily available as nitrate and therefore no adjustment was needed in regard to nitrogen.

The major identified organic constituent, p-chlorobenzene sulfonic acid was readily removed during HBC treatment. The 35% DOC that was not removed during the RBC treatment was not identified by a GS-MS scan, be- cause of the nature of the compoundicompounds.

The leachate final concentrations were 110 mg/L for DOC 1 mg/L for SBOD 370 mg/L for SCOD

These results showed close to 100% removal for biode- gradable oragnics as defined by SBOD. They also showed that a significant fraction of refractory organics remained after the treatment. In the case of DOC only 63% was removed and only 54% of the SCOD was removed.

In order to meet the limits specified for disposal into the Mill Creek Wastewater Treatment Plant, it was necessary to treat the leachate with activated powdered carbon to remove additional refractory organics. The HBC contents were transferred to a mix tank where 12 to 15 grams of activated carbon per gram of DOC were added. Separa- tion of the carbon from the leachate was achieved by flocculation with ferric chloride during clarification. The previously specified limits were met:

TOX was under 5 mg/l 0 DOC = 20 mg/1 0 SCOD = 79 mg/l The results from the treatment of leachate with an RBC,

followed by activated powdered carbon and clarification with ferric chloride is shown in Table 2.

ECONOMIC EVALUATION

The process used at Stringfellow employs regenerated activated granular carbon for removing the soluble organics from the leachate. The cost for regenerated activated granular carbon is $290,000 per year based on the following factors: A leachate treatment rate of 96 m3/d (25000 gal/d); average TOC inlet concentration of 350 mg/L; effluent TOC concentration of 30 mg/L; carbon usage is 0.35L leachate/g (41 gal/lb C); carbon dosage is 8.2 g C/g DOC. The cost ofregenerated carbon is $2.86/kg ($1.2911 b).

The savings per year that can be realized by treating the leachate with an RBC is estimated at $190,000 through savings in carbon costs. These savings are offset b y the RBC capital costs which are estimated at $630,000 based on the results obtained at the pilot plant. Nine RBC’s are required to satisfy the 4 days of reaction time determined in the pilot plant study. The estimated installed cost is $630,000. Approximately 3.3 years of operation will be required to recover the capital expenditures for the proposed RBC’s.

However the RBC is capable of removing the equivalent of 440 mg/h.m2 (40 mg/h.fP) SBOD in one hour when operating on primary effluent. When operating on Stringfellow leachate, the RBC removed only 19 mg/h.m2 (1.7 mg/h.ft2) of SBOD in one hour indicating that the biomass may only be 10% acclimated; thus causing the low reaction rate. A study on site is recommended to determine whether a day to day operation with leachate using a full scale RBC could reduce the reaction time. This would provide a biomass that was indigenous to the leachate and possibly yield a reaction rate that was significantly faster than the rate obtained during these experiments.

The biochemical kinetics for Stringfellow leachate is less than 10% of the specific reaction rate for primary effluent, and may be attributed to the reaction rates of the specific chemicals involved in the biodegradation reaction.


CAPITAL SAVINGS APPROACH SUMMARY

The scaleup for Stringfellow is based on providing sufficient time of reaction, 4 days, by installing nine standard size RBC’s, which contain 45 m3 (12000 gal) of leachate based on .0049 m3/m2 (0.12 gal/fP) of surface area, or 410 m3 (108,000 gal) total. Four days of leachate generation averages 380 m3 (100,000 gal). If we assume the worse condition, and that is, the specific reaction rate is less than 10% of the rate for PE, then a standard size RBC may have adequate biomass present on its 100,000 square feet of surface area. We can assume the oxygen transfer rate as equivalent to the rate obtained with primary effluent when the speeds of rotation are the same, 1.5 rpm, and therefore more than adequate for satisfying the leachate requirements. Then the biochemical reaction can be as- sumed to be reaction rate limited and a modification to the RBC is proposed that would decrease the number of RBC’s from nine to two. The tankage for each of the two proposed RBC’s would be increased from 45 m3 (12000 gal) to 280 m3 (75000 gal) each to take into consideration adequate reaction time, fill time, and drain time. This approach would provide enough volume to satisfy the reaction time of 4 days and yet it would contain only 9300 m2 (100,000 fp) of surface area for each of the two RBC’s. There should be sufficient disk area and oxygen transfer capability to satisfy the reaction rate required for Stringfellow leachate. This equipment modification would reduce the capital costs from $630,000 to $200,000 and cut the capital recovery time from 3.3 years to 1.1 years. This design approach increases the attractiveness of the cost savings that can be realized by installing RBC’s at Stringfellow.

The study on treating Stringfellow leachate with a RBC showed that: 65% of the dissolved organic carbon can be removed by biological treatment. The residuals remain- ing after biological treatment will require further process- ing with activated carbon to achieve levels below 100 mg DOC/L. The direct scaleup of the pilot plant result to the Stringfellow site would require 3.3 years of operation to recover the capital costs for a RBC treatment stage. A modified design scaleup based on satisfying the reaction time requirements is an attractive alternative to reduce capital costs and thus reduce the capital cost recovery period from 3.3 years to 1.1 years.

LITERATURE CITED

1. Record of Decision, “Remedial Alternative Selection,” Stringfellow Acid Pits, Glen Avon, California, U.S. EPA, Re- gion IX, San Francisco, CA.

2. “Vapor Space Organics,” In-house method developed by the Metropolitan Sewer District of Greater Cincinnati, Industrial Waste Section, Cincinnati, Ohio.

3. Internal Memo “Stringfellow Pretreatment Plant Operation and Maintenance Costs,” Camp, Dresser & McKee, Septem- ber, 1986.

4. Brenner, R. C., J . A. Heidman, E. J. Optaken, and A. C. Petrasek, “Design Information on Rotating Biological Con- tactors,” EPA-600/2-84-106, June, 1984.

5. Opatken, E. J., “An Alternative RBC Design - Second Order Kinetics,” Eno. Prog., 5, No. 1, February, 1986.

6. Antonie, R. L., Fixed Biologicul Surfuces - Wustewuter Treutment, CRC Press, 1976.

Environmental Progress (Vol. 7, No. 1) February, 1988 45

stringfellow leachate treatment with rbc

Documents